DMSP:ssj4 midnit

From CedarWiki
Jump to: navigation, search

DMSP data logo

DMSP SSJ4 Midnight Boundary Index Obtained from Precipitating Electrons



Please read the DMSP "Rules of the Road" notice before using any of these data.

The Midnight Boundary Index (or the Auroral Boundary Index, ABI) is an estimate of the equatorward boundary of precipitating auroral electrons as determined by the SSJ/4 (F06 through F15) and SSJ/5 (F16 and beyond) instruments on DMSP spacecraft. Midnight Boundary data are available in yearly ascii files below, at the CEDAR Madrigal Archive at , and at the CEDAR Database as plots and data and as files from the Virtual Solar Terrestrial Observatory at with examples described at .

In March 2009, all the files were replaced to add additional information. Previously, only the year (#1), daynumber (#2), UTsec (#3), and estimated midnight equatorward boundary in magnetic latitude (#4) were given. The additional information included the satellite ID (#5, F06=7540, F07=8541, F08=9543, F09=0542, F10=1544, F11=2546, F12=3545, F13=4547, F14=5548, F15=6549, F16=7554, F17=8551, F18=9553), where the ten's place digit indicates '4' or '5' for SSJ/4 or SSJ/5. Extra parameters were the geographic latitude times 10 (#6) and the magnetic local time (MLT) in seconds (#7) of the equatorward boundary crossing which was then estimated at the midnight MLT (#4).

In April 2011, the processing was removed from Hanscom AFB in Massachusetts to the new location of AFRL at Kirtland AFB in New Mexico. Data were missing for 22 months, and then resumed with approximately 3 times more estimates in 2011 than before and two additional parameters: (#8) an integer from 1-3 which is the quality control (QC) flag with 3 values (best=1, good=2, not good=3) where best values are post sunset and pre-dawn, and worst values are noon and midnight, and each group contains about 1/3 of the total observed equatorward boundaries; and (#9) a number from about 0.01-8.99 which is a 'provisional Kp' (pKp) estimate. Only ABI estimates with a quality code of 1 (very good) were included until 2013, which is why there was a tripling of the estimates to include quality codes 2 and 3.

In October 2013, all ABI estimates were revised to include 13 parameters, where the previous #7-9 were shifted to #10-12 positions. Three of the new parameters were the location of the observed eqb at glat/MLT (#6 and new #10) in geographic longitude (#7), magnetic latitude (#8=eqb), and magnetic longitude (#9). The last parameter (#13) was the probability that the pKp estimate falls within 1.5 sigma of the real Kp estimate. Four days of initial observations from F06 at the end of 1982 were also added at this time.

For each MLT hour, there are linear correlations of the auroral equatorward boundary latitudes with Kp. The transformation of the equatorward boundary (eqb) at each MLT hour to the equatorward boundary at midnight (ABI) is done at the same time a pseudo or provisional Kp (pKp) value is found. The QC and probability ratio are measures of how well the 'provisional Kp' deduced from the eqatorward auroral boundary at some MLT hour correlates with the real Kp. Note that pKp can be negative (unreal), even for QC=1. The standard deviation or sigma of 27 consecutive estimates of pKp from the available DMSP satellites at the observed equatorward boundaries is computed, where the time in question is the middle point (#14 of 27). The data are binned separately in the NH and SH, in 4 seasons (daynumbers 312-035, 035-127, 128-219, 220-311), 4 geographic longitudes (0-90, 90-180, 180-270, 270-360E), and 24 MLTs, for each DMSP satellite. The sigma of 27 pKp is shifted by 1 observation to calculate a new sigma over the sliding window of 27 consecutive observations. This sliding calculation of sigma is done from the beginning of the DMSP data set for F06 in 1982 to the present. The probability is the ratio in the satellite/pole/season/glon/MLT bin of the number of times this bin had a pKp within 1.5 sigma compared to the number of points total in this bin, which is in the thousands. The probability for each bin varies with each satellite, where the boundary ratios are chosen so there are approximately 33% of the pKp estimates with QC=1, 2, and 3. Typically, ratios >~0.94 are QC=1 while ratios <~0.89 are QC=3. Clearly, the biggest factor in this QC and probability is the MLT bin, as described below.

Determination of the quality depends on a variety of factors. Generally,the clearest boundaries occur premidnight, that is, on the evening side, but not prior to 17 MLT. The next best are on the dawn side. Here, however, the flux levels are decreasing and the boundaries are not sharp. The worst boundaries are on the dayside, and near midnight when the oval is cut obliquely. Therefore, the preponderance of boundaries with the 'best' flag are on the evening side. Previously, only ABI estimates flagged as 'best' were released, but now the QC flag is included and the data set is tripled in size. If the 'provisional Kp' deduced from the eqatorward auroral boundary at some MLT hour does not correlate well with the real Kp as determined by the probability, then the QC flag is set to 3. An individual pKp could still be good for QC=3, and ABI found for QC=3 are even more likely to be good, especially if the probability is relatively high for this MLT bin. HOWEVER, USERS ARE ENCOURAGED TO USE ABI ESTIMATES WITH QC=3 WITH CAUTION AND TO AVOID pKp VALUES FOR QC=3.

Midnight Boundary data are also available from AFRL upon request from If the data are used in scientific studies, they should be referred to as "The Air Force Research Laboratory Auroral Boundary Index" and acknowledged as below. These data are in the public domain and no further permission is required. Please send a courtesy copy of any publications using the ABI to Dr. Gordon.

Acknowledgements: The DMSP particle detectors were designed by Dave Hardy of the Air Force Research Laboratory, and the Auroral Boundary Indices are provided with permission from the Space Vehicle Directorate, Air Force Research Laboratory, Kirtland AFB, NM 87117, [via the Cedar Database at the National Center for Atmospheric Research which is supported by the National Science Foundation.] OR [via the CEDAR Archival Madrigal Database at Millstone Hill, MIT which is supported by the National Science Foundation.] OR [no via if from AFRL].

Start and Stop Dates: (From Ernie Holeman, revised in consultation with Barbara Emery)

  • SSJ4 F06 1982 362 1987 203
  • SSJ4 F07 1983 328 1988 117
  • SSJ4 F08 1987 176 1994 213
  • SSJ4 F09 1988 39 1992 94
  • SSJ4 F10 1990 341 1997 318
  • SSJ4 F11 1991 337 1999 161
  • SSJ4 F12 1994 246 2002 215
  • SSJ4 F13 1995 88 2009 322
  • SSJ4 F14 1997 118 2005 272
  • SSJ4 F15 1999 351 2007 334
  • SSJ5 F16 2003 247
  • SSJ5 F17 2006 315
  • SSJ5 F18 2009 296


PLOTS Jan 1994 to Mar 2006

This x-y plot shows Midnight Boundary variations during a 27 day period.

Thumbnail Plot of Midnight Boundary Index

To see plots of the midnight boundary index from January 1994 to March 2006, click on the items below. Plots were discontinued as being of limited interest to users.

Written by Frederick Rich. Point of Contact: Dr. Gordon Wilson


It has long been known that increased auroral activity, as indicated by many ionospheric and magnetospheric parameters such as brightness of auroral arcs, disturbances in the magnetic field, magnitude of the convection electric field and magnitude of field aligned currents, is accompanied by the expansion of the auroral oval.

A statistical study of the electron precipitation observed by SSJ4 instruments on DMSP spacecraft to Kp values was done by Gussenhoven et al. (1981). They determined the regression coefficients for a linear fit of equatorward auroral boundaries for 13 of the 24 hourly local time sectors and compared the results to the Kp indices. They identified equatorward auroral boundaries by hand, using precipitating electron measurements made onboard the DMSP/F2 satellite. A follow-on study (Gussenhoven et al., 1983) developed a computer algorithm for boundary selection and increased the database to 20 of 24 local time sectors by including data from DMSP F4. This work has continued to date with all available data from the SSJ4 instruments flown on DMSP spacecraft.

Although the auroral equatorward boundary, when drawn in geomagnetic coordinates, is nearly circular for all levels of auroral activity, its center is offset from the magnetic pole toward the post-midnight local time sector. The offset is a function of magnetic activity. Therefore, individual boundary measurements made for the same auroral activity, but at different local times will differ, and, as such, cannot be directly scaled to auroral activity without removing the local time variation. Removal of the local time variation is reasonably easy to accomplish by statistically determining the position of the auroral oval for every local time sector as a function of some magnetic activity index, such as Kp.


  • Gussenhoven, M.S., D.A. Hardy, and W.J. Burke, DMSP/F2 electron observations of equatorward auroral boundaries and their relationship to magnetospheric electric fields, J. Geophys. Res., 86, 768, 1981.
  • Gussenhoven, M.S., D.A. Hardy, N. Heinemann, and E. Holeman, 1978 Diffuse Auroral Boundaries and a Derived Auroral Boundary Index, AFGL-TR-82-0398, Air Force Geophysics Laboratory, Hanscom AFB, MA, (ADA130175), 1982.
  • Gussenhoven, M.S., D.A. Hardy, and N. Heinemann, Systematics of the equatorward diffuse auroral boundary, J. Geophys. Res., 88, A7, 5692-5708, 1983.
  • Hardy, D. A., and M. S. Gussenhoven, A Statistical Model of Auroral Electron Precipitation, J. Geophys. Res., 90, A5, 4229-4248, 1985.
  • Hardy, D. A., E. G. Holeman, W. J. Burke, L. C. Gentile and K. H. Bounar (2008), Probability distributions of electron precipitation at high magnetic latitudes, Journal of Geophysical Research, Volume 113, Issue A6, doi10.1029/2007JA012746.
  • Schumaker, T.L., D.A. Hardy, S. Moran, A. Huber, J. McGarity, and J. Pantazis, Precipitating lon and Electron Detectors SSJ/4) for the Block 5D/Flight 8 DMSP Satellite, AFGL-TR-884030, Air Force Geophysics Laboratory, Hanscom AFB, MA, (ADA203990) 1988.
  • Madden, D., and M. S. Gussenhoven, Auroral Boundary Index from 1983 to 1990, Tech Report GL-TR-90-0358, Air Force Geophysics Laboratory, Hanscom AFB, MA, 21 Dec. 1990.

Last Updated

Last Update 16 May 2006 by Frederick Rich. Present Point of Contact: Gordon Wilson and Ernie Holeman

Last Update: 25 October 2013 by Barbara Emery